Lecture 20: Slow Slip Events and Stress Transfer. GEOS 655 Tectonic Geodesy Jeff Freymueller

Transcription

1 Lecture 20: Slow Slip Events and Stress Transfer GEOS 655 Tectonic Geodesy Jeff Freymueller

2 Slow Slip Events From Kristine Larson

3 What is a Slow Slip Event? Slip on a fault, like in an earthquake, BUT Slow: hours to years rather than seconds Releases little high-frequency seismic energy, so it is sometimes called a type of aseismic slip. However, many or most of these events do produce a tremor-like seismic signal. Discrete events, rather than continuous creep/slip Probably frictionally controlled slip, but not unstable like seismic slip conditional stability from rate and state friction? Now known to be very common at and below base of main seismogenic zone at subduction zones.

4 Older Slow Slip Events Not generally accepted as real until ~15-20 years ago. Suggestion of a slow precursor before 1960 Chile earthquake (a few minutes prior) Slow Slip Before 1944 Tonankai, 1946 Nankaido earthquakes Survey misclosures, tide gauges, water well level changes prior to each earthquake Investigated as possible earthquake precursors Linde and Sacks (2002) showed that observations were consistent with slow slip on plate interface below main seismogenic zone. Some Silent Earthquakes identified through normal modes; some did not correspond to regular quakes

5 San Juan Bautista 1993 M. Johnston, USGS

6 Mostly Silent Earthquake Significant fault creep and near-fault strain accompanied swarm of small earthquakes Creep and strain too large to explain by earthquakes Interpretation is that a patch on the fault was creeping What triggered what?

7 Kenai Combination of locked subduction zone (NNW) postseismic deformation (SSE) Up to 55 mm/yr relative to NOAM Up to ~75 mm/yr relative motions Along-strike changes in seismogenic zone

8 Kenai Detail Obvious transition between western and eastern Peninsula Look at sites same distance from trench Edge of plate coupling toward western edge of Peninsula Edge of PWS asperity

9 Regional Plate Coupling Slip deficit/v plate Zweck et al. (2002)

10 Updated Version Suito and Freymueller (2009)

11 and velocities relative to NOAM

12 Comparison of Time Series TALKEETNA REEDY ATWC TRAIL LAKES

13 Afterslip Model for Time Series Non-linear inversion to fit each time series n(t) = a + bt + clog(1+(t-t 0 )/τ) Logarithmic decay characteristic of afterslip (Marone et al., 1991) Rate and state-dependent friction law Velocity-strengthening Subjected to sudden stress step t 0 = to τ = 0.3 to 0.6 years ( days) Hutton et al. (2001) found days for Jalisco

14 Campaign vs. Continuous

15 and velocities relative to NOAM

16 Summary of Observations Velocities over an area >15,000 km 2 changed dramatically at ~ Large southward component, decaying with time Anomalous displacement ~log(1+t/0.6) Functional form for afterslip in velocity-strengthening material obeying rate and state dependent friction law Preceded by decrease in seismicity rate within slab

17 Intepretations Most compatible with a creep-type process on the plate interface downdip of seismogenic zone NOT a transition from locked to creep, but from one rate of creep to a faster rate Trigger for event not clearly understood NO significant (M > 5.5) earthquakes NO apparent offset in time series = no sudden creep Possible link to continuing post-1964 slip transient Did postseismic creep on adjacent segment trigger faster creep on this segment? Tide gauge observations at Anchorage suggest complex creep events have occurred in the past

18 Non-linear Deformation

19 Three Time Periods Velocities measurably different over area >100x200 km 2

20 Before and After

21 Data and Model

22 Comparison of Slip Models

23 Slip Models through Time Ohta et al. (2006)

24 Slip Models through Time Ohta et al. (2006)

25 Slip Models through Time Ohta et al. (2006) The only difference between the two time periods is accelerated slip in one patch during SSE Located downdip of 1964 earthquake rupture Also associated with seismic tremor.

26 Tokai Slow Slip Event

27 Guerrero Slow Slip Events From Kristine Larson Wide variety of events from Guerrero, Mexico Variety of spatial scales, durations, magnitudes Some events propagated along strike for a considerable distance.

28 Locked Introduction to Stress Transfer How does slip change stresses in surrounding area? BEFORE Creeping

29 Locked Introduction to Stress Transfer How does slip change stresses in surrounding area? AFTER Creeping

30 Effect of Slip Slip reduces shear stress in region that slipped, increases shear stress in surrounding region Slip may also change normal stresses. Postseismic deformation also changes stresses. Stress changes from one earthquake may bring another part of the fault or another fault closer to failure triggering.

31 Stress Transfer, or Conversations between Earthquakes 1999

32 Sequence Has Repeated

33 Bay area shocks during the 75 years before 1906

34 Bay area shocks during the 75 years after 1906

35 Bay area is a system of roughly parallel faults

36 Bay area faults may have fallen under a stress shadow in 1906 from Harris & Simpson (1998) and Parsons (2003)

37

38

39 Coulomb Failure Criterion Slip on a fault will occur if the shear stress resolved on the fault plane exceeds the force of friction retarding slip: τ µ(σ p) Define the Coulomb stress change as: ΔCFS = Δτ µδ(σ p) ΔCFS = Δτ µ'δσ

40 How to Calculate Start with a model of source of deformation Usually slip on a fault using Okada s formulation of dislocation theory Calculate strain tensor at desired points by calculating strain components at a depth of choice Convert strain tensor to change in stress tensor using linear elasticity Resolve delta-stress tensor onto desired fault plane(s) based on geometry of fault

41 Effect Depends on Orientation of Receiver fault Stress tensor changes depend on the source fault Coulomb stress also depends on the geometry of receiver fault Fancy graphics for Coulomb stress change assume receiver fault Specific fault or optimally-oriented strike slip

42 from Toda & Stein (JGR, 2003)

43

44 Stress accumulated since great 1857 shock loads Coast Ranges thrust faults

45 1992 M=7.3 Landers shock increases stress at Big Bear Landers Big Bear Los Angeles from Stein (2003) First 3 hr of Landers aftershocks plotted

46 1992 M=7.3 Landers shock promotes the M=6.5 Big Bear shock 3 hr later Landers Big Bear Los Angeles from Stein (2003) First 3 hr of Landers aftershocks plotted

47 and promotes the M=7.1 Hector Mine shock 7 yr later Hector Mine Los Angeles First 7 yr of aftershocks plotted from Stein (2003)

48 North Anatolian fault Stress changes calculated for right-lateral faults paralleling the North Anatolian fault due to the entire 20th century sequence of earthquakes Each earthquake releases stress where it slips, and brings adjacent segment closer to failure

49 2003 M=6.5 San Simeon earthquake ratcheted up stress at Parkfield 2003 Calculation by Shinji Toda on 31 Mar 2004 (a similar plot by Bob Simpson appears in Hardebeck et al, 2004)

50 Stress Transfer at Subduction Zones

51 Effect of Large vs. Small Quakes Compares effect of 1960 Chile with 1995 Antofagasta earthquakes Note difference in color scale

52 Now what? å

53 Now what part 2?

54 Now what part 2?